Preview

PDF, English - main document Download (11MB) | Lizenz: Alternative pre-mRNA splicing as a source of cancer neoepitopes by Knudsen, Arne underlies the terms of Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

Abstract

Robust identification of neoepitopes is crucial for the efficacy and safety of immunotherapy, the most promising treatment strategy for several cancer types. Current approaches have provided limited numbers of immunogenic and tumor-specific targets, thus preventing the broad application of targeted immunotherapy. Here, the focus on somatic mutation-derived neoantigens often overlooks possible neoepitopes originating from mRNA processing events. A potential new source of tumor-specific peptides is alternative pre-mRNA splicing, a widely dysregulated process in several cancer subtypes. However, there is limited insight regarding the potential of alternative splicing to generate peptides that are also presented on the cell surface. Thus, in this thesis, I aimed to investigate how perturbation of the splicing machinery contributes to the neoepitope repertoire in tumor cells.

To explore alternative splicing-derived neoantigens, I performed immunopeptidomics to determine the HLA-I ligandome of wild-type RPE-1 cells and RPE-1 cell lines carrying common cancer mutations. To facilitate the presentation of alternative splicing-derived neoepitopes, I treated these cell lines with the splicing inhibitor GEX1A. I then performed HLA-I immunopurification to recover HLA-I-bound peptides of these cells, followed by peptide identification through mass spectrometry. To be able to identify non-canonical peptides from mass spectra, I generated sample-specific custom reference databases based on matching RNA-seq data. This strategy allowed me to identify more than 8,000 unique HLA-I-presented peptides per cell line. In parallel, I specifically identified neoepitopes originating from aberrant alternative splice events. By performing differential splicing analysis between the various conditions, I obtained thousands of differentially regulated splice junction events. Particularly in cells treated with the splicing inhibitor GEX1A, alternative splicing analysis revealed numerous novel, non-annotated splice events. To examine whether these dysregulated events were translated into novel peptides, I subsequently mapped the candidate peptides to the differential splice events. With this strategy, I was able to identify and validate several alternative splicing-derived neoepitope candidates that exhibited a high immunogenic potential in in vivo immunization assays.

In conclusion, my work demonstrates that pharmacological modulation of the splicing machinery has the potential to promote the presentation of neoepitopes derived from alternative splice variants. These findings have potential implications for immunotherapy of cancer types with low tumor mutational burden. Exploring the splicing-derived neopeptidome could reveal novel therapeutic targets and serve as a predictive biomarker for response to immune checkpoint blockade therapy.